PDF Loan Contracting in the Presence of Usury Limits: Evidence ...

[Pages:53]Loan Contracting in the Presence of Usury Limits:

Evidence from Automobile Lending

DRAFT 2

Brian Melzerand Aaron Schroeder,

September 25, 2015

Abstract

We study the effects of interest rate ceilings on the market for automobile loans. Binding usury limits do not prevent high-risk borrowers from receiving credit. We find instead that loan contracting and the organization of the loan market adjust to facilitate loans to risky borrowers. When usury restrictions bind, auto dealers finance a greater share of their customers' purchases and raise the vehicle sales price (and loan amount) relative to the value of the underlying collateral. By doing so, they arrange loans with similar monthly payments and price credit risk through the mark-up on the product sale rather than the loan interest rate. Meanwhile, risky borrowers that choose non-dealer loans receive lower interest rates as intended by the law, but also receive smaller loans relative to the value of the collateral. Despite having little effect on who b-melzer@kellogg.northwestern.edu; Kellogg School of Management, Northwestern University, 2001 Sheridan Road, Evanston, IL 60208 Aaron.Schroeder@; Consumer Financial Protection Bureau, 301 Howard St., Ste. 1200 San Francisco, CA 94105 The views expressed are those of the author and do not necessarily reflect those of the Consumer Financial Protection Bureau or the United States. We thank Neale Mahoney for helpful suggestions. We are also grateful for feedback received at the Consumer Financial Protection Bureau Bag Lunch, George Washington School of Business-Federal Reserve Board Financial Literacy Seminar, Kellogg Finance Bag Lunch, Midwest Finance Association 2015 Annual Meeting and Yale University-Innovations for Poverty Action Conference on Financial Inclusion. We thank Paolina Medina-Palma for research assistance.

1

receives credit, usury limits therefore have a substantial effect on who provides credit and on the terms of credit granted.

1 Introduction

Regulation shapes consumer credit markets in many ways, whether by governing the disclosure of loan terms, promoting fair access to credit, or even placing a ceiling on interest rates. In this paper, we explore how interventions of the latter type -- usury restrictions -- influence the market for automobile loans. Usury restrictions are often motivated by the argument that lenders, if unchecked, will exercise market power and raise interest rates on risky borrowers beyond the level required to compensate for credit losses. Supporters of usury limits thus argue that lenders will respond to interest rate caps by extending credit at lower prices. Opponents counter that price ceilings will cause credit rationing, which reduces access to credit and harms precisely the risky borrowers that supporters of usury limits intend to help.

We find that neither of these theories fully describes the impact of usury limits on the market for subprime auto loans. Instead, we find that auto dealers creatively contract around binding usury limits by financing their customers' purchases and pricing default risk through the mark-up on the vehicle sale rather than through the interest rate. In the resulting equilibrium, we find that few borrowers, if any, are excluded from the market, but that auto dealers provide captive financing for a larger share of auto purchases and borrowers face different loan terms--lower interest rates but larger loan-to-value ratios--than they would in the absence of usury limits.

The strategy pursued by auto dealers is simple. Auto loans are structured as installment contracts that require constant monthly payments for a fixed maturity (typically 3-6 years) and allow the lender to repossess the vehicle if the borrower defaults. Holding fixed the

2

loan maturity and principal amount, a lender is typically constrained to adjust the monthly payment by changing the interest rate specified in the contract. When prevented from charging an interest rate sufficient for a borrower's credit risk, the lender can only serve the borrower profitably by reducing risk, for example by requiring more collateral relative to the loan amount. For a lender that also serves as the vehicle seller, however, there is an additional degree of freedom. The integrated dealer-lender can design an installment contract with the desired monthly loan payment and similar risk profile by increasing the loan amount rather than the interest rate. This strategy is not feasible without integration of sales and lending, since increasing the loan amount requires the lender to provide more cash to the borrower at origination for no compensation (the monthly loan payment and collateral are unchanged)-- indeed, as discussed above, we would expect a third-party lender to reduce the loan amount relative to collateral value. The integrated dealer-lender, meanwhile, internalizes the benefit from selling the vehicle at a larger mark-up, so can subsidize a negative net present value loan with a higher-margin sale. The dealer ultimately trades the same collateral for the same monthly payment, regardless of the contractually-specified loan amount and interest rate, so is indifferent between the interest rate-constrained and the interest rate-unconstrained contract.

Drawing on these insights, we answer the following four questions. First, do usury limits prevent risky borrowers from obtaining auto loans? Second, are auto dealers more likely to provide credit when usury limits are more binding? Third, in transactions where usury limits are more likely to bind, do dealers provide loans with lower interest rates, but similar loan payments and larger loan amounts relative to the collateral value? Fourth, do non-dealer loans display the opposite pattern, with the loan-to-value ratios decreasing (and collateral protection increasing) when the usury limit binds?

Our primary analysis uses detailed data on used automobile transactions compiled by Experian. The data pair information on vehicle transactions--vehicle type, estimated value, lienholder name, dealer name and location--with information from the purchaser's credit

3

record--loan amount, payment and duration, in addition to credit score. The full sample covers approximately 28 million vehicle sales between January 2011 and August 2013. In further analysis we use the Consumer Financial Protection Bureau's (CFPB) Consumer Credit Panel, a national sample of de-identified consumer credit records.

Despite receiving relatively little attention in recent studies of consumer credit markets, usury limits matter for a significant proportion of auto loan transactions. The majority of states limit interest rates, with ceilings ranging between 17 and 36 percent. Moreover, a thriving market for auto loans exists among subprime customers, which means that many borrowers do not qualify for interest rates beneath these ceilings. In the first half of 2014, 31%, or $70.7 billion, of auto loans went to consumers with credit scores below 640 (Equifax, 2014). Finally, the strategy by which national credit card and mortgage lenders have typically avoided usury limits--chartering as a national bank exempt from state lending laws or locating in a state without usury limits such as South Dakota--proves infeasible for most auto loans. Specifically, the local dealer's involvement as a credit intermediary has led banks to follow the lending laws applicable in the dealer's state rather than their own jurisdiction.

Although usury limits do indeed bind for many borrowers, we find little evidence that these restrictions prevent households from obtaining auto loans. Comparing states with and without usury limits, we observe a similar distribution of borrowing across credit scores. The riskiest borrowers therefore continue to receive auto loans, and comprise a similar share of borrowers, regardless of whether a state imposes an interest rate ceiling. This finding holds not only in the cross-section of states, but also in the time-series for Arkansas, which raised its usury limit from below 10% to 17% in the late-2000s. The share of auto loans granted to subprime borrowers in Arkansas tracks closely with peer states both leading up to, and following, the change in usury limit.

In contrast, we find that the organization of the vehicle loan market changes quite dramatically in the presence of usury limits. Integrated dealer-lenders provide a larger proportion of loans to subprime borrowers in states with usury limits. Notably, the likelihood of dealer

4

financing is not uniformly higher in these states. Rather, dealer financing increases particularly for subprime borrowers, for whom usury limits are more likely to bind. These findings are consistent with our hypothesis that auto dealers play an important role in facilitating loans for risky borrowers that cannot receive credit from outside lenders.

Next, we examine loan contracting, beginning with dealer-financed loans. In a simple descriptive analysis of loans in Arkansas, the state with the most stringent usury limit, we show striking evidence that loan-to-value ratios adjust upward to price credit risk when interest rates are constrained. Meanwhile, monthly loan payments trend upward with credit risk but do not rise disparately in Arkansas relative to California, the largest state without a usury restriction.

To provide more comprehensive evidence on the impact of usury limits, we use regression analysis to control for heterogeneity in borrowers and vehicles. We use a two-stage regression procedure. First, we predict the likelihood that a borrower faces a binding usury ceiling, given the borrower's credit score and the usury ceiling in his state. Second, we use the first-stage predictions in a model that quantifies the difference in loan terms where the usury ceiling is predicted to bind. The identifying variation in our model comes from variation in the tightness of usury limits across states. This two-stage procedure allows us to avoid using an endogenous measure?the realized interest rate?when measuring the extent to which the usury ceiling binds.

We find that usury limits are effective in reducing interest rates, but also lead to higher loan-to-value ratios in dealer-financed loans. Our estimates imply that a binding usury limit reduces interest rates by six-to-eight percentage points, on average. Consistent with the prediction that borrowers take out larger loans and pay higher prices relative to collateral value, we find that loan-to-value ratios increase substantially, by roughly 70 percentage points, when usury limits bind. For the bottom-line price of credit, the monthly loan payment, we find no difference due to binding usury restrictions.

The final portion of our analysis examines the terms of non-dealer loans. The regres-

5

sion analysis again shows that interest rates decline when usury limits bind, by a similar magnitude as for dealer-financed loans. Loan-to-value ratios, on the other hand, change in the opposite direction as in the sample of dealer loans: we observe significantly lower loanto-value ratios when the usury limit binds. This evidence suggests that lenders strengthen collateral requirements in order to serve borrowers for whom the interest rate limit binds. The contrast in findings between the dealer and non-dealer samples fits with our predictions outlined above. Dealers can price credit risk by raising sales prices and loan amounts, whereas non-dealers must mitigate risk in order to profitably serve borrowers for whom they cannot charge sufficiently high interest rates. The contrast in findings is also relevant when interpreting the evidence on dealer loans. The elevated loan-to-value ratios that we observe among dealer loans are specific to that subset of loans, and are not symptomatic, therefore, of an omitted variable correlated with state usury restrictions.

Our paper contributes most directly to the literature on usury restrictions. Though our paper is the first to study the impact of usury restrictions on subprime auto lending, we build on a series of related papers (Adams et al. (2009); Einav et al. (2012);Einav et al. (2013)) that use data from a single auto dealer to explore the role of liquidity constraints in borrowing, the effect of credit scoring on firm profitability and the optimal design of loan contracts. This paper also relates to a larger literature on market participants' behavior under constraints, such as Attanasio et al. (2008) and Johnson and Li (2010). Attanasio et al. found, using auto loan data in the Consumer Expenditure Survey, consumers' responsiveness to different loan characteristics change with income. This matches the commonly held belief that highincome, unconstrained consumers shop on total price measures such as interest rate, while subprime consumers focus on periodic costs such as monthly payment and loan term. If consumers indeed focus on some loan outcomes more than others, this would more easily allow lenders to avoid usury limits. Similarly, Assun?ca~o et al. (2013) examines the impact of changes in expected return on assets via changes in repossession laws, and finds impacts to both access to credit and terms offered. Additionally, examination of the impacts of

6

usury cap restrictions on other loan characteristics mirrors the literature on attribute-based regulation, including Ito and Sallee (2014), which examines the impact of basing vehicle gas mileage standards on weight, and finds bunching near the limits occurring as a result. Finally, the ability of some subprime dealers to retain profits from both the financing of the vehicle as well as the overall sale price resembles many features of the captive auto financing market. That industry features in a long line of literature, such as Barron et al. (2008), which examines the changes to credit standards resulting from the bundling of durable goods with financial products.

2 The Market for Vehicle Financing

Automobile dealers play an integral role in facilitating loans for their customers. Among "prime" customers, who have relatively strong credit histories and low risk of default, dealers often arrange financing for the customer at the time of the sale.1 Although these dealers play a role in brokering and pricing loans, they rarely act as the lender and they serve few customers for whom interest rate restrictions bind.

Our analysis focuses primarily on the segment of the auto dealer market that serves "subprime" customers, for whom default risk is high and for whom interest rate restrictions may bind. Many dealers in this segment of the market do not simply arrange loans, they actually finance their customers' purchases. In industry parlance, these locations are Buy Here Pay Here (BHPH) dealers, meaning that they sell the vehicle and also collect the recurring loan payments at the dealership.

BHPH dealers are typically independent from vehicle manufacturers and sell used cars that are older and of lower value than the inventory carried by dealers serving prime customers.2 Customers do not shop for a particular vehicle and then negotiate a purchase price

1Relatively few customers -- less than 20% of those who finance purchases -- arrange financing directly with a lender prior to negotiating the final sale of the vehicle. Much more commonly, the dealer arranges financing for the customer at the time of the sale. Davis (2012) estimates roughly 80 percent of consumers who finance a vehicle purchase use this "indirect" method in which the dealer serves as an intermediary.

2In principle, our insights apply to vehicle financing of all types, including financing of new car purchases.

7

contingent on financing. Instead, BHPH transactions usually begin with loan underwriting, as the salesperson reviews the customer's credit history, current income and major expenses, and specifies the maximum monthly loan payment for which the customer qualifies. The customer then examines the vehicles for which this payment qualifies them, and the negotiation proceeds from there to find an acceptable vehicle and agree upon the down payment and loan terms. An important aspect of this sales process, from the perspective of our analysis, is that it treats the purchase and financing as a bundle. The price of the vehicle and the loan are not presented and negotiated separately.

Our study evaluates the role of usury restrictions in encouraging dealer financing. We acknowledge, however, that there are other market frictions that may encourage dealer financing among subprime buyers. Dealers may, for example, gather useful information about borrower credit risk during the sales process, giving them an underwriting advantage over outside lenders. Dealers may also have an advantage in recovering value from defaulted loans since they can avoid transactions costs and liquidate collateral through their own dealership. Our analysis does not examine these motivations for dealer financing.

2.1 Loan Contracting with Arm's Length Financing

To clarify the way in which dealer financing changes loan contracting, we offer a stylized model of the vehicle sales and financing process. Figure 1 summarizes the cash flows among the dealer, customer and lender.

The customer and dealer agree to a sales contract in which the dealer exchanges the vehicle for a price (P), to be funded at the closing through a down payment (D) from the borrower and a payment (L) from the lender to the dealer. The borrower and lender, in turn, agree to a loan contract specifying the loan amount (L, transferred to the dealer on the borrower's behalf), the schedule of promised loan payments to be paid by the borrower, and the collateral that the lender can repossess and liquidate in the event of default. The

Usury restrictions, however, have little practical impact for the vast majority of customers that buy new cars, which have higher values.

8

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download